Project Summary/Abstract Violence within families of young children is pervasive and has considerable adverse psychiatric, psychosocial, cognitive, and physical health consequences for child and adult victims and witnesses. To design methods to effectively prevent or intervene in this major public health problem, more specific and nuanced information is needed about incidents of family violence, including what factors immediately precipitate violence and promote its persistence within incidents of violence as well as individual differences in how such factors operate. The first aim of this study is to describe and compare between- and within-person rates, frequencies, and severity of incidents of parent to child violence (PCV) and intimate partner violence (IPV) as well as the trajectory by which the severity of violence increases or decreases within the course of incidents (Aim 1). Additionally, this study is uniquely designed to measure within-incident co-occurrence (or ?spillover?) of PCV and IPV. As such, the between- and within-person rates, frequencies, and severity of distinct patterns of violence spillover will be described and compared to violent incidents that occur without cross-dyad spillover, then characteristics of violence in the first family subsystem that predict the occurrence and severity of violence in a subsequent family subsystem will be tested (Aim 2). Finally, given extensive overlap between trauma exposure and family violence perpetration, diverse forms of trauma-related perceived threat (e.g., rejection, dominance) and theoretically related contextual factors will be examined as immediate, incident-level precipitants of the occurrence, persistence, and trajectory/acceleration of PCV, IPV, and spillover. To understand whether trauma exposure and global (individual-level) threat perception are translated into immediate, incident-specific predictors of violence, additive and moderating effects of trauma history and global threat perception on the link between incident-level threat perception and the persistence and acceleration of PCV, IPV, and spillover will also be tested (Aim 3). Doing so will help reconcile empirical literatures that disagree regarding the conditions under which trauma and biased threat perception promote or inhibit violent behavior. Using non-identifiable data collection procedures to minimize under-reporting of sensitive information, online protocols will measure individual factors and repeated in-depth monthly telephone interviews will measure the process and context of family violence. Because early childhood is a time of increased family stress when rates of PCV peak and children are most likely to be involved in episodes of IPV, participants will include 200 couples (i.e., 400 individuals) with a child age 3-5 years enrolled in an urban or semi-rural Head Start program. These procedures will increase feasibility (including recruitment of fathers/father figures) and the racial/ethnic diversity of the sample, reduce method variance, alleviate many limitations of traditional measures family violence, and target low-income families who are at greatest risk for family violence.